applied-cli 0.6.3__tar.gz → 0.6.5__tar.gz

{applied_cli-0.6.3 → applied_cli-0.6.5}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: applied-cli
-Version: 0.6.3
+Version: 0.6.5
 Summary: CLI and shared client library for Applied Labs AI support agents
 Author: Applied Labs
 License-Expression: MIT

{applied_cli-0.6.3 → applied_cli-0.6.5}/applied_cli/cli.py

@@ -1916,6 +1916,27 @@ def scenario_bulk_run(
     target_agent_id: str = typer.Option(
         None, "--target-agent-id", help="Optional target agent for reruns"
     ),
+    contact_email: str = typer.Option(
+        None,
+        "--contact-email",
+        help="Run as a contact with this email (resolves/creates it) so test "
+        "conversations carry an email — fixes 'Email is not present' failures",
+    ),
+    contact_id: str = typer.Option(
+        None, "--contact-id", help="Run scenarios as this existing contact UUID"
+    ),
+    anonymous: bool = typer.Option(
+        False, "--anonymous", help="Run with an anonymous contact"
+    ),
+    wait: bool = typer.Option(
+        False, "--wait", help="Poll until all runs finish, then print final status"
+    ),
+    wait_timeout: float = typer.Option(
+        300.0, "--wait-timeout", help="Max seconds to wait with --wait (default 300)"
+    ),
+    poll_interval: float = typer.Option(
+        3.0, "--poll-interval", help="Seconds between status polls with --wait"
+    ),
     shop_id: str = typer.Option(None, "--shop-id", help="Override shop ID"),
     format: str = typer.Option(
         "text", "--format", "-f", help="Output format: text or json"
@@ -1929,6 +1950,12 @@ def scenario_bulk_run(
             scenario_ids=_parse_csv_option(scenario_ids),
             benchmark_id=benchmark_id,
             target_agent_id=target_agent_id,
+            contact_email=contact_email,
+            contact_id=contact_id,
+            anonymous=anonymous,
+            wait=wait,
+            wait_timeout=wait_timeout,
+            poll_interval=poll_interval,
             output_format=format,
         )
     )

{applied_cli-0.6.3 → applied_cli-0.6.5}/applied_cli/tools.py

@@ -8,6 +8,7 @@ import asyncio
 import difflib
 import json
 import re
+import time
 from contextlib import suppress
 from html.parser import HTMLParser
 from typing import Any
@@ -6082,26 +6083,116 @@ async def scenario_run_delete(
     return f"Scenario run {run_id} deleted successfully."
 
 
+def _bulk_status_counts(counts: dict | None) -> dict[str, int]:
+    """Normalize bulk-status counts to lowercase keys with int values."""
+    lowered: dict[str, int] = {}
+    for key, value in (counts or {}).items():
+        lowered[str(key).lower()] = int(value or 0)
+    return lowered
+
+
+def _bulk_pending_count(counts: dict | None) -> int:
+    """Count runs still queued or running (case-insensitive)."""
+    normalized = _bulk_status_counts(counts)
+    return normalized.get("queued", 0) + normalized.get("running", 0)
+
+
+async def _await_bulk_run(
+    client: AppliedClient,
+    job_id: str,
+    *,
+    timeout: float,
+    poll_interval: float,
+) -> tuple[dict, bool]:
+    """Poll a bulk run until no runs are queued/running or the timeout elapses.
+
+    Returns (latest_status_payload, timed_out).
+    """
+    start = time.monotonic()
+    status = await client.get_scenario_bulk_run_status(job_id)
+    while _bulk_pending_count(status.get("counts")) > 0:
+        if time.monotonic() - start >= timeout:
+            return status, True
+        await asyncio.sleep(poll_interval)
+        status = await client.get_scenario_bulk_run_status(job_id)
+    return status, False
+
+
+async def _resolve_contact_override(
+    client: AppliedClient,
+    *,
+    contact_override: dict | None,
+    contact_id: str | None,
+    contact_email: str | None,
+    anonymous: bool,
+) -> dict | None:
+    """Build the contact_override payload from the convenience arguments.
+
+    An explicit contact_override dict wins. Otherwise anonymous > contact_id >
+    contact_email (which resolves/creates a contact so the test conversation
+    carries a real email).
+    """
+    if contact_override:
+        return contact_override
+    if anonymous:
+        return {"mode": "anonymous"}
+    if contact_id:
+        return {"mode": "contact", "contact_id": contact_id}
+    if contact_email:
+        contact = await client.get_or_create_contact(email=contact_email)
+        resolved_id = contact.get("id")
+        if not resolved_id:
+            raise AppliedAPIError(
+                f"Could not resolve a contact for email {contact_email}.",
+                status_code=404,
+            )
+        return {"mode": "contact", "contact_id": str(resolved_id)}
+    return None
+
+
 async def scenario_bulk_run(
     client: AppliedClient,
     scenario_ids: list[str] | None = None,
     benchmark_id: str | None = None,
     target_agent_id: str | None = None,
     contact_override: dict | None = None,
+    contact_id: str | None = None,
+    contact_email: str | None = None,
+    anonymous: bool = False,
+    wait: bool = False,
+    wait_timeout: float = 300.0,
+    poll_interval: float = 3.0,
     output_format: str = "text",
 ) -> str:
     """
     Run multiple scenarios at once.
 
+    By default a scenario run reuses the input conversation's contact, which on
+    test/benchmark conversations often has no email — causing agents to respond
+    "Email is not present in the conversation." Pass contact_email or contact_id
+    to run the scenarios as a contact that has an email, so the test conversation
+    carries it.
+
+    With wait=True, this polls until every run finishes (or the timeout elapses)
+    and returns the final status, so you can run a benchmark and read results in
+    one call instead of polling scenario_bulk_status yourself.
+
     Args:
         client: Authenticated AppliedClient
         scenario_ids: List of scenario UUIDs to run
         benchmark_id: Run all scenarios in this benchmark
         target_agent_id: Optional agent to run against (for A/B testing)
-        contact_override: Optional contact override, e.g. {"mode": "contact", "contact_id": "<uuid>"}
+        contact_override: Raw override, e.g. {"mode": "contact", "contact_id": "<uuid>"}
+            (takes precedence over the convenience args below)
+        contact_id: Run scenarios as this existing contact (gives test convos its email)
+        contact_email: Resolve/create a contact with this email and run as them
+        anonymous: Run with an anonymous contact (mode='anonymous')
+        wait: Poll until all runs finish (or wait_timeout elapses)
+        wait_timeout: Max seconds to wait when wait=True (default 300)
+        poll_interval: Seconds between status polls when wait=True (default 3)
 
     Returns:
-        Summary of runs created
+        Summary of runs created (plus the final status when wait=True)
     """
     resolved_scenario_ids = list(scenario_ids or [])
     if not resolved_scenario_ids:
@@ -6123,10 +6214,17 @@ async def scenario_bulk_run(
             )
 
     try:
+        effective_override = await _resolve_contact_override(
+            client,
+            contact_override=contact_override,
+            contact_id=contact_id,
+            contact_email=contact_email,
+            anonymous=anonymous,
+        )
         result = await client.bulk_run_scenarios(
             scenario_ids=resolved_scenario_ids,
             target_agent_id=target_agent_id,
-            contact_override=contact_override,
+            contact_override=effective_override,
         )
     except AppliedAPIError as e:
         return _format_error(e)
@@ -6144,6 +6242,26 @@ async def scenario_bulk_run(
         "contact_override": result.get("contact_override"),
     }
 
+    job_id = payload.get("job_id")
+    final_status: dict | None = None
+    timed_out = False
+    if wait and job_id:
+        try:
+            final_status, timed_out = await _await_bulk_run(
+                client,
+                str(job_id),
+                timeout=wait_timeout,
+                poll_interval=poll_interval,
+            )
+        except AppliedAPIError as e:
+            return _format_error(e)
+        counts = _bulk_status_counts(final_status.get("counts"))
+        payload["final_counts"] = counts
+        payload["timed_out"] = timed_out
+        payload["duration_seconds"] = final_status.get("duration_seconds")
+        payload["completed_at"] = final_status.get("completed_at")
+        payload["failed"] = final_status.get("failed") or []
+
     if output_format == "json":
         return to_json(payload)
 
@@ -6160,6 +6278,23 @@ async def scenario_bulk_run(
         output += f"scenario_run_ids: {preview_ids}\n"
         if len(run_ids) > 10:
             output += f"more_runs: {len(run_ids) - 10}\n"
+
+    if final_status is not None:
+        counts = payload["final_counts"]
+        output += "\n# Final Status\n"
+        output += "timed_out: " + ("true (still pending)" if timed_out else "false") + "\n"
+        output += f"completed: {counts.get('completed', 0)}\n"
+        output += f"failed: {counts.get('failed', 0)}\n"
+        pending = counts.get("queued", 0) + counts.get("running", 0)
+        output += f"still_pending: {pending}\n"
+        if payload.get("duration_seconds") is not None:
+            output += f"duration_seconds: {payload['duration_seconds']}\n"
+        failed_runs = payload.get("failed") or []
+        if failed_runs:
+            output += f"\n# Failed Runs ({len(failed_runs)})\n"
+            output += to_json(failed_runs)
+        return output
+
     output += "\nTip: use scenario_bulk_status(job_id, include_runs=True) or scenario_run_list(bulk_job_id=job_id) to get per-run details with scenario mappings."
     return output
 
@@ -6193,14 +6328,14 @@ async def scenario_bulk_status(
             payload.pop("runs", None)
         return to_json(payload)
 
-    counts = result.get("counts") or {}
+    counts = _bulk_status_counts(result.get("counts"))
     output = "# Bulk Run Status\n"
     output += f"job_id: {result.get('job_id')}\n"
     output += f"total: {result.get('total')}\n"
-    output += f"queued: {counts.get('QUEUED', 0)}\n"
-    output += f"running: {counts.get('RUNNING', 0)}\n"
-    output += f"completed: {counts.get('COMPLETED', 0)}\n"
-    output += f"failed: {counts.get('FAILED', 0)}\n"
+    output += f"queued: {counts.get('queued', 0)}\n"
+    output += f"running: {counts.get('running', 0)}\n"
+    output += f"completed: {counts.get('completed', 0)}\n"
+    output += f"failed: {counts.get('failed', 0)}\n"
     output += f"created_at: {result.get('created_at')}\n"
     output += f"updated_at: {result.get('updated_at')}\n"
     if result.get("completed_at"):

{applied_cli-0.6.3 → applied_cli-0.6.5}/applied_cli/v2/scenarios.py

@@ -27,6 +27,12 @@ class ScenariosBulkRunInput(StrictInput):
     benchmark_id: str | None = None
     target_agent_id: str | None = None
     contact_override: dict[str, Any] | None = None
+    contact_id: str | None = None
+    contact_email: str | None = None
+    anonymous: bool = False
+    wait: bool = False
+    wait_timeout: float = 300.0
+    poll_interval: float = 3.0
 
 
 class ScenariosBulkCancelInput(StrictInput):
@@ -796,10 +802,19 @@ async def scenarios_bulk_run_handler(
             )
 
     try:
+        from applied_cli.tools import _resolve_contact_override
+
+        effective_override = await _resolve_contact_override(
+            client,
+            contact_override=params.contact_override,
+            contact_id=params.contact_id,
+            contact_email=params.contact_email,
+            anonymous=params.anonymous,
+        )
         result = await client.bulk_run_scenarios(
             scenario_ids=resolved_scenario_ids,
             target_agent_id=params.target_agent_id,
-            contact_override=params.contact_override,
+            contact_override=effective_override,
         )
     except AppliedAPIError as exc:
         return _api_error_result(exc)
@@ -816,6 +831,48 @@ async def scenarios_bulk_run_handler(
         "duplicated_scenarios": result.get("duplicated_scenarios"),
         "contact_override": result.get("contact_override"),
     }
+
+    job_id = payload.get("job_id")
+    if params.wait and job_id:
+        from applied_cli.tools import _await_bulk_run, _bulk_status_counts
+
+        try:
+            final_status, timed_out = await _await_bulk_run(
+                client,
+                str(job_id),
+                timeout=params.wait_timeout,
+                poll_interval=params.poll_interval,
+            )
+        except AppliedAPIError as exc:
+            return _api_error_result(exc)
+        counts = _bulk_status_counts(final_status.get("counts"))
+        payload["final_counts"] = counts
+        payload["timed_out"] = timed_out
+        payload["duration_seconds"] = final_status.get("duration_seconds")
+        payload["failed"] = final_status.get("failed") or []
+        pending = counts.get("queued", 0) + counts.get("running", 0)
+        summary = (
+            f"Bulk job {job_id} "
+            + ("timed out with " if timed_out else "finished: ")
+            + f"{counts.get('completed', 0)} completed, "
+            + f"{counts.get('failed', 0)} failed"
+            + (f", {pending} still pending" if pending else "")
+            + "."
+        )
+        warnings = []
+        if counts.get("failed"):
+            warnings.append(f"{counts['failed']} run(s) failed.")
+        if timed_out:
+            warnings.append("Timed out before all runs finished.")
+        return ToolResult(
+            data=payload,
+            summary=summary,
+            warnings=warnings,
+            next_actions=[
+                "Use scenarios_bulk_status with include_runs=true to inspect runs.",
+            ],
+        )
+
     queued = payload.get("queued") or 0
     return ToolResult(
         data=payload,
@@ -1050,7 +1107,10 @@ def scenario_specs() -> list[ToolSpec]:
             namespace="scenarios",
             description=(
                 "Run selected scenarios or every scenario in a benchmark and "
-                "return the queued job metadata."
+                "return the queued job metadata. Pass contact_email or contact_id "
+                "to run as a contact with an email (fixes 'Email is not present' "
+                "failures on test conversations). Pass wait=true to block until "
+                "all runs finish and return the final status in one call."
             ),
             input_model=ScenariosBulkRunInput,
             output_model=None,

{applied_cli-0.6.3 → applied_cli-0.6.5}/applied_cli.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: applied-cli
-Version: 0.6.3
+Version: 0.6.5
 Summary: CLI and shared client library for Applied Labs AI support agents
 Author: Applied Labs
 License-Expression: MIT

{applied_cli-0.6.3 → applied_cli-0.6.5}/applied_cli.egg-info/SOURCES.txt

@@ -50,6 +50,8 @@ tests/test_flow_tools.py
 tests/test_knowledge_content_tools.py
 tests/test_recovery.py
 tests/test_scenario_bulk_cancel.py
+tests/test_scenario_bulk_run_contact.py
+tests/test_scenario_bulk_run_wait.py
 tests/test_toolkit_contract.py
 tests/test_v2_agents.py
 tests/test_v2_articles.py

{applied_cli-0.6.3 → applied_cli-0.6.5}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "applied-cli"
-version = "0.6.3"
+version = "0.6.5"
 description = "CLI and shared client library for Applied Labs AI support agents"
 readme = "README.md"
 requires-python = ">=3.11"

applied_cli-0.6.5/tests/test_scenario_bulk_run_contact.py

@@ -0,0 +1,116 @@
+import json
+
+import pytest
+
+from applied_cli import tools
+
+
+class FakeRunClient:
+    def __init__(self):
+        self.bulk_kwargs = None
+        self.get_or_create_calls = []
+
+    async def list_scenarios(self, benchmark_id=None, limit=500, **kwargs):
+        return [{"id": "s1"}, {"id": "s2"}]
+
+    async def get_or_create_contact(self, email=None, name=None, phone=None):
+        self.get_or_create_calls.append(email)
+        return {"id": "contact-123", "email": email}
+
+    async def bulk_run_scenarios(
+        self, scenario_ids=None, target_agent_id=None, contact_override=None
+    ):
+        self.bulk_kwargs = {
+            "scenario_ids": scenario_ids,
+            "target_agent_id": target_agent_id,
+            "contact_override": contact_override,
+        }
+        return {
+            "job_id": "job-1",
+            "total": len(scenario_ids or []),
+            "queued": len(scenario_ids or []),
+            "scenario_run_ids": ["r1", "r2"],
+            "contact_override": contact_override,
+        }
+
+
+@pytest.mark.asyncio
+async def test_no_contact_args_sends_no_override():
+    client = FakeRunClient()
+    await tools.scenario_bulk_run(client, benchmark_id="bench-1", output_format="json")
+    assert client.bulk_kwargs["contact_override"] is None
+
+
+@pytest.mark.asyncio
+async def test_contact_id_builds_override():
+    client = FakeRunClient()
+    await tools.scenario_bulk_run(
+        client, benchmark_id="bench-1", contact_id="c-9", output_format="json"
+    )
+    assert client.bulk_kwargs["contact_override"] == {
+        "mode": "contact",
+        "contact_id": "c-9",
+    }
+
+
+@pytest.mark.asyncio
+async def test_contact_email_resolves_then_overrides():
+    client = FakeRunClient()
+    result = await tools.scenario_bulk_run(
+        client,
+        benchmark_id="bench-1",
+        contact_email="casey@example.com",
+        output_format="json",
+    )
+    assert client.get_or_create_calls == ["casey@example.com"]
+    assert client.bulk_kwargs["contact_override"] == {
+        "mode": "contact",
+        "contact_id": "contact-123",
+    }
+    assert json.loads(result)["job_id"] == "job-1"
+
+
+@pytest.mark.asyncio
+async def test_anonymous_mode():
+    client = FakeRunClient()
+    await tools.scenario_bulk_run(
+        client, benchmark_id="bench-1", anonymous=True, output_format="json"
+    )
+    assert client.bulk_kwargs["contact_override"] == {"mode": "anonymous"}
+
+
+@pytest.mark.asyncio
+async def test_explicit_override_wins_over_convenience_args():
+    client = FakeRunClient()
+    await tools.scenario_bulk_run(
+        client,
+        benchmark_id="bench-1",
+        contact_override={"mode": "contact", "contact_id": "raw"},
+        contact_email="ignored@example.com",
+        output_format="json",
+    )
+    # The raw override is used; email resolution is skipped.
+    assert client.get_or_create_calls == []
+    assert client.bulk_kwargs["contact_override"] == {
+        "mode": "contact",
+        "contact_id": "raw",
+    }
+
+
+@pytest.mark.asyncio
+async def test_v2_handler_threads_contact_email():
+    from applied_cli.v2.scenarios import (
+        ScenariosBulkRunInput,
+        scenarios_bulk_run_handler,
+    )
+
+    client = FakeRunClient()
+    result = await scenarios_bulk_run_handler(
+        client,
+        ScenariosBulkRunInput(benchmark_id="bench-1", contact_email="x@example.com"),
+    )
+    assert client.bulk_kwargs["contact_override"] == {
+        "mode": "contact",
+        "contact_id": "contact-123",
+    }
+    assert result.data["job_id"] == "job-1"

applied_cli-0.6.5/tests/test_scenario_bulk_run_wait.py

@@ -0,0 +1,107 @@
+import json
+
+import pytest
+
+from applied_cli import tools
+
+
+class FakeWaitClient:
+    """Bulk client whose status transitions to done after N polls."""
+
+    def __init__(self, status_sequence):
+        self._status_sequence = list(status_sequence)
+        self._poll = 0
+        self.status_calls = 0
+
+    async def list_scenarios(self, benchmark_id=None, limit=500, **kwargs):
+        return [{"id": "s1"}, {"id": "s2"}]
+
+    async def bulk_run_scenarios(
+        self, scenario_ids=None, target_agent_id=None, contact_override=None
+    ):
+        return {
+            "job_id": "job-1",
+            "total": len(scenario_ids or []),
+            "queued": len(scenario_ids or []),
+            "scenario_run_ids": ["r1", "r2"],
+        }
+
+    async def get_scenario_bulk_run_status(self, job_id):
+        self.status_calls += 1
+        idx = min(self._poll, len(self._status_sequence) - 1)
+        self._poll += 1
+        return self._status_sequence[idx]
+
+
+@pytest.mark.asyncio
+async def test_wait_polls_until_no_pending(monkeypatch):
+    # Avoid real sleeping between polls.
+    async def _no_sleep(_seconds):
+        return None
+
+    monkeypatch.setattr(tools.asyncio, "sleep", _no_sleep)
+
+    client = FakeWaitClient(
+        status_sequence=[
+            {"counts": {"queued": 2, "running": 0, "completed": 0, "failed": 0}},
+            {"counts": {"queued": 0, "running": 1, "completed": 1, "failed": 0}},
+            {
+                "counts": {"queued": 0, "running": 0, "completed": 2, "failed": 0},
+                "duration_seconds": 12.5,
+                "completed_at": "2026-06-05T10:00:00Z",
+                "failed": [],
+            },
+        ]
+    )
+    result = await tools.scenario_bulk_run(
+        client, benchmark_id="bench-1", wait=True, output_format="json"
+    )
+    data = json.loads(result)
+    assert data["timed_out"] is False
+    assert data["final_counts"]["completed"] == 2
+    assert data["duration_seconds"] == 12.5
+    assert client.status_calls == 3
+
+
+@pytest.mark.asyncio
+async def test_wait_times_out_when_runs_stay_pending(monkeypatch):
+    async def _no_sleep(_seconds):
+        return None
+
+    monkeypatch.setattr(tools.asyncio, "sleep", _no_sleep)
+
+    # Always pending → must hit the timeout path.
+    client = FakeWaitClient(
+        status_sequence=[
+            {"counts": {"queued": 2, "running": 0, "completed": 0, "failed": 0}}
+        ]
+    )
+    result = await tools.scenario_bulk_run(
+        client,
+        benchmark_id="bench-1",
+        wait=True,
+        wait_timeout=0.0,  # immediate timeout after first poll
+        output_format="json",
+    )
+    data = json.loads(result)
+    assert data["timed_out"] is True
+    assert data["final_counts"]["queued"] == 2
+
+
+@pytest.mark.asyncio
+async def test_no_wait_returns_started_summary():
+    client = FakeWaitClient(status_sequence=[{"counts": {}}])
+    result = await tools.scenario_bulk_run(
+        client, benchmark_id="bench-1", output_format="json"
+    )
+    data = json.loads(result)
+    assert "final_counts" not in data
+    assert client.status_calls == 0  # no polling when wait is False
+
+
+def test_bulk_status_counts_normalizes_case_and_types():
+    assert tools._bulk_status_counts({"QUEUED": 2, "Running": "1"}) == {
+        "queued": 2,
+        "running": 1,
+    }
+    assert tools._bulk_pending_count({"QUEUED": 3, "RUNNING": 4, "COMPLETED": 9}) == 7